Introduction

Top Abstract Introduction Results Discussion References

Irinotecan (7-ethyl-10-[4-(1-piperidino)-1-piperidino]carbonyloxycamptothecin; CPT-11¹) (Fig. 1) is a semisynthetic water-soluble derivative of camptothecin, with a heterocyclic ring structure isolated from the Chinese tree Camptotheca acuminata (Sawada et al., 1991). CPT-11 has been shown to have strong antitumor activity through inhibition of topoisomerase I (Creemers et al., 1994; Pommier et al., 1994). This drug is currently registered for use in Japan, Europe, and the United States in patients with metastatic colorectal cancer refractory to several first-line therapies, including 5-fluorouracil, and shows clinical activity against several other types of solid tumors (Slichenmyer et al., 1993; Wiseman and Markham, 1996).

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Chemical structures of CPT-11 and SN-38.

After intravenous administration, CPT-11 is metabolized by carboxylesterases in the liver and small intestine to an active metabolite, 7-ethyl-10-hydroxycamptothecin (SN-38) (Fig. 1), which is 100- to 1000-fold more potent than CPT-11 as a topoisomerase I inhibitor in various in vitro system (Creemers et al., 1994; Pommier et al., 1994; Mathijssen et al., 2001). SN-38 is further conjugated by UDP-glucuronosyltransferases primarily in liver to an inactive -glucuronide derivative and excreted into urine and bile (Rivory, 2000; Mathijssen et al., 2001). Hydrolysis of SN-38 glucuronide by the intestinal microflora can occur and allows possible recycling of SN-38 in humans (Rivory, 2000; Mathijssen et al., 2001). However, CPT-11 unexpectedly causes severe toxicity of leukopenia or diarrhea, and it has been suggested that these side effects are due to an accumulation of SN-38 in various tissues (Rothenberg, 1997; Mathijssen et al., 2001).

Another pathway of CPT-11 metabolism consists of a cytochrome P450 (P450)-mediated oxidation of the bipiperidine side chain attached to the core structure (Haaz et al., 1998a,b; Rivory, 2000). The main metabolites resulting from this pathway have been identified as the aminopentane carboxylic acid and primary amine derivatives of CPT-11, namely APC and NPC, respectively (Rivory, 2000; Santos et al., 2000; Slatter et al., 2000). Although these metabolites have been regarded to have little cytotoxicity like CPT-11, NPC but not APC can be converted into SN-38 by hepatic carboxylesterases and, as such, may contribute to the overall production of the pharmacologically active species (Rivory, 2000). P450 consists of a superfamily of heme-containing monooxygenases and is responsible for the metabolism of many drugs, environmental chemicals, and endogenous substances (Nelson et al., 1996). Although the number of individual isoforms that have been identified and characterized is increasing, the drug metabolism in humans is handled mainly by CYP1, CYP2, and CYP3 (Spatzenegger and Jaeger, 1995; Nelson et al., 1996). In the case of CPT-11, CYP3A4 has been suggested to be the predominant isoform in oxidative metabolism (Haaz et al., 1998a,b; Santos et al., 2000). By contrast, the products and isoforms involved in the P450-mediated metabolism of SN-38, the active metabolite of CPT-11, in humans have not been identified, although the UDP-glucuronosyltransferase isoforms involved in glucuronidation have (Iyer et al., 1998; Rivory, 2000; Mathijssen et al., 2001).

Clinically relevant drug interactions are often the result of the effects of P450 enzymes during metabolism and can cause severe complications (Muck, 1994). However, there is little information available about the clinical interactions of CPT-11 with other drugs, and there are no reports directly comparing the efficiencies of CPT-11 and SN-38 to interfere with the biological functions of P450 isoforms in humans. In this study to evaluate the potential for in vivo drug interactions involving CPT-11, the inhibition and mechanism-based inactivation potencies of CPT-11 and SN-38 for human P450 isoforms were examined using microsomes from insect cells expressing human P450s.

Experimental Procedures

Materials. CPT-11 (lot 115122) and SN-38 (lot 970326) were kindly supplied by Yakult Honsha Co. (Tokyo, Japan). The purity (99%) of each compound was confirmed by analytical HPLC. Phenacetin, 4-acetamidophenol, 3-acetamidophenol, N-phenylanthranilic acid, chlorzoxazone, and 4-nitrophenol were purchased from Aldrich Chemical Co. (Milwaukee, WI). Coumarin, 7-hydroxycoumarin, 7-ethoxycoumarin, corticosterone, glutathione, and deferoxamine were purchased from Sigma Chemical Co. (St. Louis, MO). Paclitaxel, diclofenac, and testosterone were obtained from Wako Pure Chemical Industries (Osaka, Japan). 6-Hydroxypaclitaxel and 4'-hydroxydiclofenac were obtained from GENTEST (Woburn, MA). S-(+)-Mephenytoin, (±)-4'-hydroxymephenytoin, (±)-bufuralol, (±)-1'-hydroxybufuralol, 6-hydroxychlorzoxazone, and 6-hydroxytestosterone were purchased from Ultrafine Chemicals and Research (Manchester, UK). 7,13-Diacetyl baccatin III was obtained from Alexis Biochemicals (San Diego, CA), phenobarbital from Tokyo Kasei Kogyo Co. (Tokyo, Japan), and catalase from Calbiochem-Novabiochem Co. (San Diego, CA). NADP⁺, glucose 6-phosphate, and glucose-6-phosphate dehydrogenase were purchased from Oriental Yeast Co. (Tokyo, Japan). All other chemicals and organic solvents were of the highest quality commercially available. Microsomes from baculovirus-infected insect cells expressing CYP1A2, CYP2A6 + b₅, CYP2B6 + b₅, CYP2C8 + b₅, CYP2C9*1 + b₅, CYP2C19 + b₅, CYP2D6*1, CYP2E1 + b₅, and CYP3A4 + b₅ were obtained from GENTEST. All enzymes were coexpressed with NADPH-cytochrome P450 oxidoreductase. The P450 contents were as described in the data sheets provided by the manufacturer.

Enzyme Assays. P450-dependent activities were determined by HPLC, as described previously with some modifications (Chauret et al., 1997; Busby et al., 1999; Nakajima et al., 1999). The standard incubation mixture contained 50 mM potassium phosphate buffer, pH 7.4, substrate, human P450-expressing microsomes, inhibitor/inactivator, and NADPH-generating system (1.0 mM NADP⁺, 10 mM glucose 6-phosphate, 2.0 units/ml glucose-6-phosphate dehydrogenase, and 5.0 mM MgCl₂) in a final volume of 500 µl. All substrates were dissolved in methanol. CPT-11 and SN-38 as inhibitors/inactivators were dissolved in dimethyl sulfoxide, with the solvent being used as the control. The final concentration of organic solvent (methanol and dimethyl sulfoxide) in the incubation mixture was 0.8% (v/v). The substrate concentrations for the determination of residual P450 activities were set up to be close to the K_m value for each enzyme activity obtained in the preliminary kinetic analysis. Substrate and microsomal P450 concentrations for the assays were as follows: CYP1A2, 20 µM phenacetin and 20 pmol of P450/ml; CYP2A6, 0.80 µM coumarin and 5.0 pmol of P450/ml; CYP2B6, 100 µM 7-ethoxycoumarin and 20 pmol of P450/ml; CYP2C8, 5.0 µM paclitaxel and 20 pmol of P450/ml; CYP2C9, 2.0 µM diclofenac and 10 pmol/ml P450; CYP2C19, 20 µM S-mephenytoin and 20 pmol/ml; CYP2D6, 1.5 µM bufuralol and 5.0 pmol of P450/ml; CYP2E1, 600 µM chlorzoxazone and 50 pmol of P450/ml; CYP3A4, 50 µM testosterone and 10 pmol of P450/ml. The substrate concentrations for the determination of K_i were 0.20 to 3.2 µM for CYP2A6, 0.50 to 8.0 µM for CYP2C9, and 10 to 160 µM for CYP3A4.

HPLC Analysis. HPLC analysis was performed using a Shimadzu SCL-10A system controller (Kyoto, Japan) consisting of three LC-10AD pumps, an SIL-10A auto injector with sample cooler, an SPD-10AV UV-VIS detector, a CTO-10A column oven, a DGU-14A degasser, and a C-R7A chromatopac integrator. The samples were cooled at 4°C, and 20-µl aliquots were injected into an Inertsil ODS-80A column (150 × 4.6-mm i.d.; GL Sciences, Tokyo, Japan). The column was kept at 40°C. The calibration curves were established using authentic metabolites. The analytical conditions under which the products did not overlap with interfering peaks originating from inhibitors/inactivators and their metabolites were examined in a preliminary study.

Inhibition Experiments. P450-dependent activities in the absence and presence of CPT-11 or SN-38 were measured according to the method described in the previous sections. The inhibitor concentrations were set at 100 µM. For the determination of the inhibition type and K_i values, Lineweaver-Burk plots were constructed. Michaelis-Menten parameters (K_m and V_max values) were estimated using EnzymeKinetics v.1.4 (Trinity Software, Campton, NH). The K_i values were calculated from the following equations (Voet and Voet, 1990): for competitive inhibition, K_i = K_m[I]/(K_m'  K_m), and for mixed inhibition, K_i = K_mV_max'[I]/(K_m'V_max  K_mV_max'), where [I] is the inhibitor concentration, and K_m' and V_max' are the Michaelis-Menten parameters in the presence of inhibitor.

Inactivation Experiments. Human P450-expressing microsomes were preincubated at 37°C for 20 min with CPT-11 or SN-38 at 100 µM in the presence of the NADPH-generating system. After preincubation, typical substrates were added, and the corresponding P450 activities were measured according to the method described in the previous sections. For the determination of k_inact and K_I values, human P450-expressing microsomes were preincubated at 37°C in the presence of the NADPH-generating system with varying concentrations (0, 25, 50, and 100 µM) of CPT-11 or SN-38. At selected time intervals (0, 10, 20, 40, and 60 min for CYP2A6 and CYP2C9 and 0, 5, 10, 20, and 30 min for CYP3A4), aliquots (100 µl) were withdrawn and diluted 5-fold into a prewarmed (37°C) incubation system containing NADPH-generating system and typical substrates in 50 mM potassium phosphate buffer, pH 7.4, and the corresponding P450 activities were measured according to the method described in the previous sections. The k_obs (initial rate constant for inactivation) values were obtained as slopes of an initial linear phase plotting logarithm of the residual activity against the preincubation time (Waley, 1985). The k_inact and K_I values were estimated by a nonlinear regression analysis using Prism v.3.0a (GraphPad Software, San Diego, CA). The effects of trapping agents on the inactivation of CYP3A4 were determined by incubating 100 µM CPT-11 or SN-38 together with glutathione (2.0 mM), deferoxamine (100 µM), or catalase (1000 units/ml) at 37°C for 20 min in the preactivation step.

Statistics. All reported values are the mean ± S.D. of three separate experiments in duplicate. Similar results were obtained in all experiments.

	Results

Top Abstract Introduction Results Discussion References

Inhibition of Human P450 Activities by CPT-11 and SN-38. The inhibitory effects of CPT-11 and SN-38 on the activity of each P450 isoform were examined (Fig. 2). Although we attempted to determine the IC₅₀ values of CPT-11 and SN-38, no 50% inhibitory effect up to a concentration of 100 µM was observed for any P450 isoform (data not shown). Additionally, the solubility of SN-38 in the reaction mixture was very low (<120 µM). Therefore, CPT-11 and SN-38 were used at a single concentration (100 µM) in this study. Both CPT-11 and SN-38 effectively inhibited CYP3A4 activity by 28 and 30%, respectively. SN-38 also inhibited CYP2A6 and CYP2C9 activities by 27 and 31%, although the inhibition potency was not as strong as that for CYP3A4 activity. However, CYP1A2, CYP2B6, CYP2C8, CYP2C19, CYP2D6, and CYP2E1 activities were unaffected by either compound.

View larger version (39K): [in this window] [in a new window]   Fig. 2.   Inhibition of human P450 activities by CPT-11 and SN-38. Experimental conditions are described under Experimental Procedures. Concentrations of CPT-11 and SN-38 were 100 µM. Control activities (picomoles per minute per picomole of P450) were: CYP1A2, 8.76 ± 0.12; CYP2A6, 4.29 ± 0.10; CYP2B6, 3.69 ± 0.17; CYP2C8, 2.14 ± 0.10; CYP2C9, 11.1 ± 0.3; CYP2C19, 3.07 ± 0.08; CYP2D6, 12.7 ± 0.5; CYP2E1, 6.75 ± 0.16; and CYP3A4, 49.2 ± 1.4. Each bar represents the mean of three separate experiments performed in duplicate. , CPT-11; , SN-38.

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Kinetics for inhibition of human CYP3A4 activity by CPT-11 and SN-38. Experimental conditions are described under Experimental Procedures. Each point represents the mean of three separate experiments performed in duplicate. , 0 µM CPT-11 or SN-38; , 100 µM CPT-11 or SN-38.

Mechanism-Based Inactivation of Human P450 Enzymes by CPT-11 and SN-38. The possibility that CPT-11 and SN-38 are mechanism-based inactivators of human P450 enzymes was examined (Fig. 4). CYP2A6 and CYP2C9 activities were decreased 27 and 29% by SN-38, respectively. CYP3A4 activity was notably decreased by both CPT-11 and SN-38 (60 and 82%, respectively). The efficiency of SN-38 against CYP3A4 activity was greater than that of CPT-11. The other P450 enzymes were not affected by either compound.

View larger version (38K): [in this window] [in a new window]   Fig. 4.   Inactivation of human P450 enzymes by CPT-11 and SN-38. Experimental conditions are described under Experimental Procedures. Concentrations of CPT-11 and SN-38 were 100 µM. Control activities (picomoles per minute per picomole of P450) were: CYP1A2, 5.25 ± 0.29; CYP2A6, 4.63 ± 0.13; CYP2B6, 2.37 ± 0.14; CYP2C8, 1.78 ± 0.04; CYP2C9, 9.60 ± 0.49; CYP2C19, 2.50 ± 0.07; CYP2D6, 8.26 ± 0.31; CYP2E1, 6.23 ± 0.15; and CYP3A4, 36.3 ± 0.8. Each bar represents the mean of three separate experiments performed in duplicate. , CPT-11; , SN-38.

1

1

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Time- and concentration-dependent inactivation of human CYP3A4 by CPT-11 and SN-38. Experimental conditions are described under Experimental Procedures. Each point represents the mean of three separate experiments performed in duplicate. , 0 µM CPT-11 or SN-38; , 25 µM CPT-11 or SN-38; , 50 µM CPT-11 or SN-38; , 100 µM CPT-11 or SN-38.

	Discussion

Top Abstract Introduction Results Discussion References

Although CPT-11 is a useful anticancer agent for numerous malignancies, such as colon and lung cancers, it can cause severe leukopenia or diarrhea due to an accumulation of SN-38, the active metabolite of CPT-11, in various tissues (Rothenberg, 1997). CPT-11 is clinically administered either alone or in combination with other drugs, including anticancer agents. However, coadministration often causes drug interactions leading to harmful side effects (Muck, 1994). Thus, the prediction or evaluation of the potential for drug interactions is an important aspect of individualized drug therapy. One of the most important mechanisms of drug interactions is the decrease of hepatic metabolism catalyzed by P450 enzymes (Guengerich, 1997). In the present study, the selectivity of inhibition or mechanism-based inactivation of human P450 enzymes by CPT-11 and SN-38 was investigated using microsomes from insect cells expressing P450s. To this end, we used nine P450-dependent activities as markers of P450 isoforms that mainly catalyze xenobiotics. The P450-dependent activities were phenacetin O-deethylation for CYP1A2, coumarin 7-hydroxylation for CYP2A6, 7-ethoxycoumarin O-deethylation for CYP2B6, paclitaxel 6-hydroxylation for CYP2C8, diclofenac 4'-hydroxylation for CYP2C9, S-mephenytoin 4'-hydroxylation for CYP2C19, bufuralol 1'-hydroxylation for CYP2D6, chlorzoxazone 6-hydroxylation for CYP2E1, and testosterone 6-hydroxylation for CYP3A4 (Wrighton and Stevens, 1992; Wrighton et al., 1993; Busby et al., 1999).

It has been reported that the most predominant P450 isoform involved in the oxidative metabolism of CPT-11 is CYP3A4 (Haaz et al., 1998a,b; Santos et al., 2000). The present results showed that CPT-11 effectively inhibits only CYP3A4 activity among the P450 isoforms tested, through competitive inhibition. SN-38, however, inhibited CYP2A6, CYP2C9, and CYP3A4 activities with a mixed or competitive type of inhibition. These observations may suggest that not only CYP3A4 but also CYP2A6 and CYP2C9 contribute to SN-38 metabolism. However, the inhibitory effects of CPT-11 and SN-38 on these P450 activities were not so strong; the K_i values (121-181 µM) were higher than the K_m values (18-111 µM) for the formation of APC and NPC in P450-mediated metabolism of CPT-11 using microsomes from human livers and mammalian cells expressing CYP3A4 in previous reports (Haaz et al., 1998a,b; Santos et al., 2000). On the other hand, Haaz et al. (1998a,b) have found that the oxidative metabolism of CPT-11 in human liver microsomes is remarkably inhibited by drugs such as loperamide and ondansetron, which are mainly metabolized by CYP3A4. The reason for the discrepancy between our results and previous reports may be a difference in the marker enzyme reactions, source of the enzyme, and/or CYP3A4 properties. Like CYP3A4, CYP2A6 and CYP2C9 also catalyze many drugs used clinically, such as tegafur, warfarin, and phenytoin (Rendic and Di Carlo, 1997). To predict drug interactions involving CPT-11, therefore, it is necessary to identify the rate and metabolites of the oxidative metabolism of SN-38 in humans and the P450 isoforms that are involved.

Drug metabolism catalyzed by P450 enzymes can be inhibited by other mechanisms in addition to competitive inhibition. One example is the inactivation of P450 by the metabolite of a drug that covalently binds to the enzyme to form a complex with P450, leading to irreversible inhibition (Silverman, 1988). In this case, as the drug has to be metabolically activated by an enzyme and covalently binds to the same enzyme, inactivation affects only the P450 isoform that is involved in the drug metabolism. However, it has not been determined whether CPT-11 and/or its metabolites are further activated by P450 enzymes to form active metabolites that inhibit P450-dependent drug oxidation in humans. Therefore, we also examined the possibility that CPT-11 and SN-38 are mechanism-based inactivators of human P450 enzymes. It was clearly shown that CYP3A4 is inactivated by CPT-11 and SN-38 in an NADPH-, time-, and concentration-dependent manner. The inactivation by SN-38 was more extensive than that by CPT-11. In addition, glutathione, deferoxamine, and catalase did not prevent or slow the inactivation of P450 enzymes, supporting this possibility, because a lack of prevention by these agents is one of the characteristics of mechanism-based inactivation. The inactivation kinetics of human CYP3A4 by mechanism-based inactivators have been reported previously. For example, the values of k_inact and K_I have been estimated as 0.39 min¹ and 46 µM for gestodene (Guengerich, 1990), 1.62 min¹ and 7.5 µM for L-754,394 (human immunodeficiency virus-1 protease inhibitor) (Chiba et al., 1995), 0.59 min¹ and 22 µM for delavirdine (Voorman et al., 1998), 0.09 min¹ and 4.7 µM for mifepristone (He et al., 1999), 0.06 to 0.17 min¹ and 16 to 19 µM for erythromycin (Kanamitsu et al., 2000), 0.06 min¹ and 13 µM for amiodarone (Ohyama et al., 2000), and 0.07 min¹ and 3.3 µM for diltiazem (Yeo and Yeo, 2001), respectively. The k_inact values of CPT-11 and SN-38 for CYP3A4 activities obtained in this study were comparable to those of mifepristone, erythromycin, amiodarone, and diltiazem, whereas the values were much lower than those of gestodene, L-754,394, and delavirdine. CYP2A6 and CYP2C9 were not inactivated by SN-38, although they were subject to mixed-type inhibition. This may imply that the decrease in CYP2A6 and CYP2C9 activities is not due to the formation of a metabolic intermediate complex. Accordingly, it is possible that CPT-11 and SN-38 are moderate mechanism-based inactivators of human CYP3A4.

The mechanism-based inactivation requires more attention than competitive inhibition because the inhibitory effects remain after the elimination of the inhibitor from blood and tissue. It has been reported that the terminal elimination half-life of CPT-11 in plasma after intravenous administration is 14 to 15 h, and the value of SN-38 increases 1.7- to 2.1-fold relative to CPT-11 (Sparreboom et al., 1998; Kehrer et al., 2000; Slatter et al., 2000). Furthermore, the excretion rate of SN-38 into urine and bile has been reported to be clearly slower than that of unchanged CPT-11, although the total urinary and fecal excretion as free/conjugated forms is 6.5 to 12% of the dose (Sparreboom et al., 1998; Slatter et al., 2000). These findings may mean that SN-38 is extensively distributed and accumulated in the tissues and that the metabolite rather than parent drug causes drug interactions. However, although the K_I values of CPT-11 and SN-38 for CYP3A4 activities obtained in this study differed from the plasma concentration of CPT-11 and its metabolites (0.17-7.5 µM) (Sparreboom et al., 1998; Kehrer et al., 2000), we could not identify the relationship between in vitro and in vivo data.

In conclusion, we studied the selectivity of the inhibition or inactivation of human P450 isoforms by CPT-11 and its active metabolite SN-38. CPT-11 and SN-38 blocked CYP3A4 activity and CYP2A6, CYP2C9, and CYP3A4 activities, respectively, through competitive or mixed-type inhibition. The K_i values were >100 µM, suggesting that the drug interactions involving CPT-11 caused by competitive inhibition are clinically insignificant. However, a moderate mechanism-based inactivation of CYP3A4 by CPT-11 and SN-38 was observed (SN-38 > CPT-11). Therefore, our findings imply that significant drug interactions involving CPT-11 results from a mechanism-based inactivation of CYP3A4 by SN-38 as an active metabolite of CPT-11 rather than competitive inhibition. The pharmacokinetic interactions of CPT-11 based on the induction of CYP3A4 by anticonvulsants and micellae encapsulation of anticancer agents in formulation vehicles have been reported (Friedman et al., 1999; van Zuylen et al., 2001). Further studies are underway in our laboratory to clarify the relationship between the metabolic profile of CPT-11 and the inhibition or inactivation selectivity of P450 isoforms and to predict in vivo interactions of CPT-11 from in vitro data for the suitable use of CPT-11 in combination with other drugs.